Update README.md
Browse files
README.md
CHANGED
|
@@ -14,10 +14,6 @@ tags:
|
|
| 14 |
|
| 15 |
# Model Card for Model ID
|
| 16 |
|
| 17 |
-
|
| 18 |
-
## Model Details
|
| 19 |
-
|
| 20 |
-
|
| 21 |
### Model Description
|
| 22 |
|
| 23 |
This model is a fine-tuned version of Qwen2.5-1.5B-Instruct on an Azerbaijani medical reasoning dataset.
|
|
@@ -61,7 +57,6 @@ Performance in zero-shot generalisation (e.g., English → Azerbaijani medical t
|
|
| 61 |
## How to Get Started with the Model
|
| 62 |
|
| 63 |
```python
|
| 64 |
-
|
| 65 |
login(token="")
|
| 66 |
|
| 67 |
tokenizer = AutoTokenizer.from_pretrained("unsloth/Qwen2.5-1.5B-Instruct",)
|
|
@@ -87,7 +82,6 @@ outputs = model.generate(
|
|
| 87 |
)
|
| 88 |
|
| 89 |
print(tokenizer.decode(outputs[0]))
|
| 90 |
-
|
| 91 |
```
|
| 92 |
|
| 93 |
## Training Details
|
|
@@ -97,7 +91,8 @@ print(tokenizer.decode(outputs[0]))
|
|
| 97 |
The model was fine-tuned on a translated and cleaned version of FreedomIntelligence/medical-o1-reasoning-SFT, which was manually converted into Azerbaijani.
|
| 98 |
All examples were filtered for translation quality and medical relevance.
|
| 99 |
|
| 100 |
-
Dataset(Translated): Rustamshry/medical_o1_reasoning_SFT_az
|
|
|
|
| 101 |
Link of Original Dataset: huggingface.co/datasets/FreedomIntelligence/medical-o1-reasoning-SFT
|
| 102 |
|
| 103 |
|
|
@@ -118,7 +113,7 @@ The model was trained using supervised fine-tuning (SFT) with parameter-efficien
|
|
| 118 |
- **Target Modules:** q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj
|
| 119 |
|
| 120 |
|
| 121 |
-
#### Speeds, Sizes, Times
|
| 122 |
|
| 123 |
- **Training speed:** 0.12 steps/sec
|
| 124 |
- **Total training time:** 11 hours, 26 minutes
|
|
|
|
| 14 |
|
| 15 |
# Model Card for Model ID
|
| 16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
### Model Description
|
| 18 |
|
| 19 |
This model is a fine-tuned version of Qwen2.5-1.5B-Instruct on an Azerbaijani medical reasoning dataset.
|
|
|
|
| 57 |
## How to Get Started with the Model
|
| 58 |
|
| 59 |
```python
|
|
|
|
| 60 |
login(token="")
|
| 61 |
|
| 62 |
tokenizer = AutoTokenizer.from_pretrained("unsloth/Qwen2.5-1.5B-Instruct",)
|
|
|
|
| 82 |
)
|
| 83 |
|
| 84 |
print(tokenizer.decode(outputs[0]))
|
|
|
|
| 85 |
```
|
| 86 |
|
| 87 |
## Training Details
|
|
|
|
| 91 |
The model was fine-tuned on a translated and cleaned version of FreedomIntelligence/medical-o1-reasoning-SFT, which was manually converted into Azerbaijani.
|
| 92 |
All examples were filtered for translation quality and medical relevance.
|
| 93 |
|
| 94 |
+
Dataset (Translated): Rustamshry/medical_o1_reasoning_SFT_az
|
| 95 |
+
|
| 96 |
Link of Original Dataset: huggingface.co/datasets/FreedomIntelligence/medical-o1-reasoning-SFT
|
| 97 |
|
| 98 |
|
|
|
|
| 113 |
- **Target Modules:** q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj
|
| 114 |
|
| 115 |
|
| 116 |
+
#### Speeds, Sizes, Times
|
| 117 |
|
| 118 |
- **Training speed:** 0.12 steps/sec
|
| 119 |
- **Total training time:** 11 hours, 26 minutes
|